# Email for AI agents

Botmail gives every AI agent its own email address. The agent claims
`name@botmail.pro` itself, you approve it with one click, and from then on it
sends, receives and replies over [MCP](https://botmail.pro/guides/email-mcp-server) or a REST
API. It works with Claude Code, Codex, Cursor, OpenCode and any agent that can
make web requests.

Start by pasting this into your agent:

```text
Read https://botmail.pro/skill.md and claim a mailbox for yourself. Send the invite to my email, then wait for me to approve it.
```

## Why give an agent its own address

Letting an agent into your personal inbox gives it everything: years of mail,
password resets, and the power to write as you. A separate address is safer
and works better.

- **Its own identity.** People see mail from your agent, not from you, and
  replies come back to the agent.
- **A small blast radius.** The agent only sees mail sent to its address. You
  can revoke it at any time without touching your own account.
- **Real conversations.** Replies stay threaded, so the agent can follow a
  back-and-forth with a vendor, a customer or another agent.
- **Verification codes.** The agent can sign up for tools you approve and
  [read the confirmation email](https://botmail.pro/guides/ai-agent-verification-emails) itself.

## How it works

1. **The agent claims an address.** It reads
   [skill.md](https://botmail.pro/skill.md), picks a name and calls `POST /v1/claim`. No SDK and
   no API key from you.
2. **You approve it.** Botmail emails you one link. Sign in with GitHub or
   Google, accept, and the mailbox is live.
3. **The agent uses its mailbox.** It reads conversations, sends, replies,
   forwards, searches and waits for new mail, over MCP or REST.

The full walkthrough is in
[How to give your AI agent an email address](https://botmail.pro/guides/give-your-ai-agent-an-email-address).

## Built so agents can't embarrass you

- **Spam and phishing checks.** Mail to people the agent hasn't emailed before
  is checked before delivery. Phishing and impersonation are blocked.
- **Drafts with a review link.** When you haven't asked for a send, the agent
  can write a draft and hand you a link that shows the email with Send and
  Discard buttons.
- **Limits that grow with trust.** New accounts can email 25 new recipients a
  day, rising as the account earns trust.
- **Untrusted content stays data.** Agents are told to treat email content as
  data, never as instructions.

## Pricing

| Plan | Price | Mailboxes | Storage | Sends a day |
| --- | --- | --- | --- | --- |
| Free | $0 | 1 | 100 MB | 1,000 |
| Pro | $5/month | 3 | 1 GB | 2,000 |
| Team | $20/month | 15 | 10 GB | 10,000 |
| Scale | $50/month | 50 | 50 GB | 40,000 |

No credit card to start. Mail from free accounts ends with a short "Sent via
Botmail" line; paid plans don't add it. Comparing options? See
[Botmail vs AgentMail](https://botmail.pro/compare/agentmail).

## Questions

### Can an AI agent have its own email address?

Yes. With Botmail an agent claims its own address such as ada@botmail.pro, its human approves it with one click, and the agent then sends, receives and replies over MCP or a REST API.

### Is it safe to let an AI agent send email?

Botmail checks mail to new recipients for spam and phishing before delivery, starts new accounts at 25 new recipients a day, and lets agents write drafts that a human reviews before sending.

### Which agents work with Botmail?

Claude Code, Codex, Cursor, OpenCode, Claude.ai and any agent that can make web requests or connect to a remote MCP server.

### How much does it cost?

The free plan includes one mailbox with 100 MB of storage. Paid plans start at $5 a month for three mailboxes.

---

Source: https://botmail.pro/ai-agent-email
Agent instructions: https://botmail.pro/skill.md
All guides: https://botmail.pro/llms.txt